As more and more business-critical applications are being deployed on virtualized, shared infrastructure, the number, type, and/or variety of storage devices has increased, as has the potential physical distance between storage devices belonging to the same infrastructure. A clustered storage system may include, for example, two or more cluster nodes connected in a distributed architecture such that the cluster nodes operate in a coordinated manner to respond to host requests. Each cluster node may, however, comprise different hardware, different software or firmware (or different version of the same software or firmware), and may be located physically distant from each other (in, for example, different buildings, cities, states, or countries). A need exists for a system and method for identifying and diagnosing device failures in virtualized, shared infrastructure.